An inverted U-shaped association was observed between noise levels and sleep duration, with both too little and too much noise associated with shorter sleep, and this relationship was attenuated in cities with higher green space coverage and larger population.
Key Findings
Results
Noise exposure showed an inverted U-shaped (non-linear) relationship with sleep duration, meaning moderate noise was associated with longer sleep while both very low and very high noise were associated with shorter sleep.
Linear term: β = 0.746, P < 0.001; quadratic term: β = -0.955, P < 0.001
The inflection point (peak sleep duration) occurred at a Baidu Noise Index value of 0.391 (95% CI: 0.320, 0.566)
Fixed-effects regression models were used to estimate the impact of noise on sleep duration
The finding challenges the assumption that quieter environments are always better for sleep
Results
The inverted U-shaped relationship between noise and sleep duration was attenuated in cities with higher green space coverage.
Quadratic interaction term for green space: β = 0.822, P < 0.05
Green space coverage data were obtained from the National Bureau of Statistics of China
Moderation analysis was applied to explore the role of green space
Higher green space coverage flattened the inverted U-shaped curve, suggesting green space buffers the extremes of noise's effect on sleep
Results
The inverted U-shaped relationship between noise and sleep duration was attenuated in larger cities (by population size).
Quadratic interaction term for city population: β = 1.072, P < 0.001
Urban population data were obtained from the National Bureau of Statistics of China
City size moderated the noise-sleep relationship more strongly than green space coverage based on the magnitude of the interaction coefficient
The attenuation suggests residents of larger cities may adapt differently to noise environments
Methods
City-level noise exposure was quantified using the Baidu Noise Index, described by the authors as a novel approach in this research context.
The study used data from the China Health and Retirement Longitudinal Study (CHARLS)
The analysis was longitudinal in design, using fixed-effects regression models
Fixed-effects models were employed to control for time-invariant individual-level confounders
The Baidu Noise Index provided city-level noise exposure estimates
Background
The study was conducted in the context of rapid urbanization in China, where noise pollution has become a growing public health concern.
The authors frame noise pollution as a consequence of rapid urbanization in China
The study provides findings intended to inform urban planning and health policy formulation
The authors highlight the relevance of optimizing urban sound environments and promoting green space development
The study adds to public health literature by identifying a non-linear rather than simply negative relationship between noise and sleep
What This Means
This research suggests that the relationship between environmental noise and how long people sleep is not straightforward. Rather than finding that more noise always means less sleep, the study found an inverted U-shape: people in areas with moderate noise levels tended to sleep the most, while those in both very quiet and very noisy environments slept less. The turning point occurred at a specific noise index value, beyond which additional noise was associated with declining sleep duration. The study used data from a large Chinese longitudinal health survey combined with city-level noise measurements from a digital mapping tool called the Baidu Noise Index.
The study also found that this noise-sleep relationship was influenced by two city-level factors: green space coverage and city population size. In cities with more parks and vegetation, and in larger cities, the inverted U-shape was less pronounced — meaning the extremes of noise had a smaller impact on sleep in those places. This suggests that urban greenery and the characteristics of large cities (such as better soundproofing in buildings, or residents who have adapted to noise) may buffer the effects of noise on sleep.
This research matters for urban planning and public health policy because it challenges the idea that making cities as quiet as possible is always best for sleep. It suggests that some level of ambient noise may actually be associated with better sleep — possibly because very quiet environments can be unfamiliar or even unsettling — and that investing in urban green spaces could help protect residents' sleep health regardless of noise levels.
Chen Y, Chang L, Huang Z, Lou N, Chu Y, Tu J, et al.. (2026). Is Too Noisy as Bad as Too Quiet? The Inverted U-shaped Relationship between Noise and Sleep Duration.. Noise & health. https://doi.org/10.4103/nah.nah_43_25